Astra robot model learns tasks from failures in 3 tries, dubbed robotics' ChatGPT moment
YuXiang_IRVL · x · 2026-09-15
A robotics researcher shares another Astra failure-learning demo: the model went from figuring out controls and the environment to completing a task in 3 tries, showing strong in-context learning by reasoning through mistakes (adjusting grip angle and carry height). It's slow, but the author argues Astra could be robotics' ChatGPT moment.
More from Embodied
- Open-source companion robot Lamp starts shipping to 20+ countries at $969 — dee_hw · 2026-10-30
- SCAMPER demo: virtual rat trots and gallops with gaits it never saw in training — tyrell_turing · 2026-10-30
- DIY humanoid build, day 7: creator gives up fixing physics, adds 1-DOF ankles — _Stocko_ · 2026-10-30
- Lambda Surgical launches first-in-human study of LiDAR-based LambdaGuide at PINE — ditzikow · 2026-10-30
- Gaussian splats double as shared GPS: one city map localizes drones, robots and phones — bilawalsidhu · 2026-10-30
- SimReady dataset offers 2,677 simulation-ready 3D assets to build virtual test worlds — victormustar · 2026-10-30